«We don’t need any more Linux distributions. Stop making distributions and create applications«. Lighter, the water. The phrase is not mine, but of Alan Pope, well-known developer of Canonical and Ubuntu; but as if it were, because I fully subscribe to it. And I explain, in addition to explaining the context in which that phrase occurs.
Make a Linux App
Make a Linux App is an initiative promoted by Pope in order to encourage idle developers to contribute to the GNU / Linux ecosystem in the right direction, which is not to create a new distribution that will only interest four cats – usually also idle – and that nobody is asking for. Why? Because we already have distros to spare and very good. We don’t need more!
What Pope proposes in Make a Linux App, on the other hand, is drawer: creating an application, a good application, is more complicated, but much more beneficial for all GNU / Linux users regardless of the distribution they use, because Although the lack of applications is not the weakest link in the Linux desktop experience, we are still behind Windows and Mac.
The Make a Linux App explains the reasons why creating applications for Linux is positive both for the ecosystem and for the developer himself; Starting points are recommended, for example the framework to use, including that of GNOME, KDE, elementary OS, Ubuntu Touch and (bravo to point out, despite the opposition it has among many users) Electron; as well as all the current distribution possibilities are remembered, from AppImage to Flatpak, Snap and the OpenSUSE Open Build Service.
How could it be otherwise, Pope has been fair enough not to highlight Canonical’s solutions above the rest. The initiative also has the support of the main stakeholders, read projects such as GNOME, KDE, elementary or UBPorts. Take a look at the site; It is worth it and all the information is concise, it will not make you dizzy with a thousand data.
The unbearable lightness of some Linux distributions
Now, why should the developer working on his own distribution, usually rehash of the rehash, take this initiative? Because directly and indirectly, is what most users are asking for. As it is. We have a good example of this in the popularity indices of distros, which I will illustrate with the results of our year-end survey.
If you look, of the 20 options available, the bulk of the votes will go to the big ones: Ubuntu, Linux Mint, Debian, Manjaro, Arch Linux, KDE neon, Fedora, Deepin, elementary OS, openSUSE … The popularity of the rest is residual and its user base will go hand in hand, with few exceptions. But even if we add them together, many other small distros that do not add value continue to be left out.
To give a couple of examples that you will know and whose contribution is more substantial, in recent times we have recommended MX Linux, a derivative of Debian that I consider quite interesting. However, the most interesting thing is not the distro itself, but the tools it provides. Wouldn’t it be possible to abstract those tools from the distribution and offer them as a software package for Debian? The same for Peppermint, which as its main value has to offer a tool for creating web applications.
Eye: they are two examples taken on the fly and as I have said, their contribution is more substantial than others, in which their existential justification is based on modifying the appearance and adding tons of pre-installed software. Also, it is not always easier for developers and users to add packages than to use something ready from the very installation. However, even in these two cases we are talking about residual distros that, if disappeared, would not affect the Linux desktop at all.
Distros as a tool vs general-purpose distros
On the other hand, we must not confuse what are distributions that work as a tool, to general purpose distributions, that is, those that we install on our PC. Ubuntu, Fedora or Manjaro have nothing to do with Kali, Tails, Puppy, Robolinux, LibreELEC, SteamOS, SystemRescueCd … And so many others, whose focus is not to serve as an operating system for day to day, but meet a specific need. This is one of the great things about GNU / Linux that shouldn’t change.
Redundancy does not equate to fragmentation
Nor should redundancy be confused with fragmentation. There is no noticeable fragmentation in GNU / Linux, although there are distributions that contribute little or nothing and the reason I have already pointed out: the big ones eat almost all the cake. Referring again to the results of our end-of-year survey, everything based on Debian took 70% of the votes … but the fact is that of that percentage, more than 50% was distributed among Ubuntu, Linux Mint and Debian ; the rest went for KDE neon, Deepin, elementary OS, MX Linux, Zorin OS and stop counting.
Then the main inconvenience of creating more general-purpose distributions is not that it generates fragmentation or affects the ecosystem: it is that it does not benefit it. It’s that simple. Of course…
Free software is for good and for “bad”
Here is the crux of the matter, and that is that free software is free for everything, also for those who wish to set up their own distribution, instead of contributing to other areas where effort is more welcome. As the famous Linus Torvals once said, free or open source software has triumphed on the basis of the selfish kind of “I do this because it benefits me”, with the exception that the fruit is shared.
Therefore, it is enough that there is someone who wants to do their rehash for them to do it, no one can stop them. What we can do, and that’s what Make a Linux App is about, is ask you to consider other forms of contribution. How to create an application or, perhaps, help maintain one that needs it. And they are many.